ML19211C543

Forwards NRC marked-up & Describing Concerns Re Incompleteness of Audit Calculations on BWRs & Pwrs.Lists Two Types of Audit Calculations That Should Be Performed

ML19211C543
Person / Time
Issue date:	08/08/1979
From:	Thadani A NRC - TMI-2 UNRESOLVED SAFETY ISSUES TASK FORCE
To:	Aycock M NRC - TMI-2 UNRESOLVED SAFETY ISSUES TASK FORCE
Shared Package
ML19211C544	List: ML19211C543
References
REF-GTECI-A-09, REF-GTECI-SY, TASK-A-09, TASK-A-9, TASK-OR NUDOCS 8001110658
Download: ML19211C543 (6)

Text

,. - -

0 o

UNITED STATES

~

84' NUCLEAR REGULATORY COMMISSION g

5 E

WASHINGTON, D, C. 20555 i

f o

%**v /

AUG 8 1979 NOTE T0: d B. Aycock, Deputy Director Unresolved Safety Issues Program FR0ft:

A. Thadani, Unresolved Safety Issues Program Enclosures 1 and 2 describe some of my concerns on the incompleteness of our audit calculations on BWRs and PWRs respectively.

I see a need for a short term (1 s 2 months) as well as long tem (3 s 4 months) effort to conduct some audit calculations to confim our past judgments on ATWS. The manpower needs and the computer time estimates are preliminary and were provided by M. Levine of BNL.

BWRs All ATWS calculations to date performed by GE have utilized "REDY" code.

Some audit calculations were performed by BNL in 1973, 1974.

Subsequent tests at the Peach Bottom reactor indicated some inadequacies of the REDY code. Currently GE uses a 1.D "0DYN" code for all overpressure transient events.

The staff is adamant that Turbine Trip Without Bypass (TTWOBP)

C WS overpressure event be analyzed using "0DYN" code. As discussed in, two types of audit calculations should be performed.

Type 1:

Short Term Plant Response Analyze two ATWS transients, TTWOBP and MSIV closure Carry calculations up to 1 minute real time Codes:

TWIGL - RELAP-3B Manpower:

2 men - 4 to 6 weeks Computer Time:

5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> Type 2:

Long Term Plant Response (s 10 min.)

Analyze effects of Boron injection on plant response for TTWOBP and MSIV closure ATWS events.

Codes:

RELAP-38 tbnpower:

1 man - 2 months Computer Time:

4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> 1736 230 soon a.y

M. B. Aycock AUG B 1979 PWRs As explained in Enclosure 2, the staff audit calculations addressed only over-pressure concern and not the potential for core uncovering for some ATWS events. Thus there is a need for the following audit calculations (20 m 30 minutes real time) for each vendor design.

Transients: Loss of feedwater with stuck open valve Loss of offsite power with stuck open valve a.

base case 95% MTC, HEM b.

99% NTC c.

Time delay in aux feed and 1/2 aux feed d.

0.9 HEM e.

HPSI Design effects Codes:

IRT - RELAP-3B Manpower: 2 man - 4 months Computer Time: 80 hours9.259259e-4 days <br />0.0222 hours <br />1.322751e-4 weeks <br />3.044e-5 months <br /> Comment:

If we expect to go to Commission by December 1979, then we need as a minimum short term BWR audit calculation as well as the PWR calculations.

If the PWR calculations show serious core uncovering problem, then I think we should discuss that with the Commission before December 1979 because our per-ception of higher risk from BWRs may not be quite correct.

Because BNL staff is committed on other tasks (some physics type manpower may be available), J spoke with Richard Denning and Bob Collier of BCL and they indicated their knowledge of RELAP and their willingness to provide personnel to go to BNL and use BNL facilities to perform these tasks.

BNL is receptive to the idea of getting this help from BCL. Thus with coordinated effort between NRC/BNL/

BCL we may be able to meet the following schedule if we begin work by 9/1/79.

BWRs Type 1:

Complete by 10/30/79 Type 2: Complete by 11/30/79 PWRs Preliminary Assessment 11/30/79 Studies Complete 2/28/80 Total Manpower s 13 Man Months Total Computer Time s 90 hours0.00104 days <br />0.025 hours <br />1.488095e-4 weeks <br />3.4245e-5 months <br /> 1736 231

M. B. Aycock Al'C 8 1979 The need to perform BWR ATWS analysis:

Short Term:

Previous audit calculations were performed using point kinetics. The model did not include steam line dynamics. The previous GE code used for ATWS analysis is the REDY code. The REDY code did not predict Peach Bottom test results where steam line dynamics and space kinetics were important. The REDY code predicted neutron flux peak nonconservatively by a factor of 2 to 3.

The need for fairly accurate heat flux cannot be overemphasized because of the resul-tant effects on containment and other structures.

This is particularly im-portant for plants with alternative #3 fix. On a best estimate basis the REDY code is not acceptable for sudden overpressurization transients, General Electric submitted the ODYN code for the analysis of these transient and they do not seek the approval of REDY for sudden overpressurization transients.

Most ATWS events result in rapid overpressure condition. Hence, previous analyses performed by GE should be reperformed using the ODYN code at least to verify previous analyses. The staff should reperform the audit calcula-tions using steam line dynamics and space kinetics models.

Long Tenn:

Audit calculations were not performed to verify GE calculations for long tera behavior. The effectiveness of the fixes " Boron reactivity feedback" was never evaluated.

Both short and long term energy releases are important to evaluate torus behavior.

The frequency and duration of the opening of the valves are governed by the effectiveness of the fix and the dy(namics of the steamline, Because of the criticality of alternative #3 fix small margin to limit) and its impact on consequences, it is necessary to perfonn some audit calculations.

The need to perfonn PWR ATWS analyses:

TMI-2 event showed that some transients may lead to boiling in the primary system loop and eventually to core uncovering. The previous ATWS analyses were performed evaluating the overpressurization effects which occur for a short time in the beginning of the transient. The aspects of core uncover-ing and boiling in the primary system were overlooked.

It is necessary to establish:

1) the validity of the vendor codes used in the ATWS analysis if there is some boiling and 2) if there is boiling, does the core uncover, We need audit calculations to answer these questions.

1736 232

M. B. Aycock AUG 8 1979 TMI-2 event also showed that ATWS audit calculations must cover n me failures which impact consequences.

These failures are 1) stuck open val ve, 2) delay in auxiliary feedwater and 3) reduction in auxiliary feedwater. We need audit calculations to establish sensitivities of the ATWS events to these failures.

Further an inadvertent opening of a safety or relief valve is an anticipated transient which may have significant consequences and audit cal-culations are necessary to confirm vendor analysis.

(Note vendor analyses are probably inadequate).

The potentially serious consequences for some design (different HPSI shut off head) should be carefully reviewed and audit calculations of such cases are warranted.

< (';

ft ' 0A?

A. Thadani Unresolved Safety Issues Program cc: ATWS Distribution

Enclosure:

As statea 1736 233

$N(l.0,$'/ff

/

.,/

NUCLEAR REGULATCRY CG,. /,iS3 TON 5 kl-.[. 5 v.mNcio. o c. nsss

.e Q ^ -d /

....a c,.:.. _ 1. _., -:. ::

Generic Task Action A-9 D"

D P D

$n R

b e

NOTE T0:

S. H. Hanauer, Director Unresolved Safety Issues Program FROM:

Ashok Thadani

SUBJECT:

BWR CODE EULUATI0!i Most of the ATWS analyses were perforred by GE using their code called "REDY" as described in NEDD-lCB02.

Our audit calculations using RELAP only cover several seconds (

25 seconds) of an ATWS transient and the calculated peak pressures agreed well with the GE calculational result.

How ver. GE is roi using "0DYt;" code for ali ove pressure events.

This r.ew code..as de. eloped by GE because the Peach Bot Or turbine tric test results did re: a c r e e '.. ell with the results p edi::e: ty '- sEDY '.

Tr e' 5;:11:stic.n of OT. ccdi to A c:S events has not yet mn r eviered by the staff.

I see a need for tne folic.c-ing effort in code s aluation.

1.

Review "0DYN" f -

~.0

pli:=tien 2.

Perform Audit Calcu'.a:icas a) First several :e:c-ds of ATWS event - deter

.r.e flux, p n re.

S/R discharge (tne n.cdel includes RPT).

b) Several. +

.c' P'.:S E cer:

.:se H ~N

a_CS inje:: ion rates. injr.: :n -i e anc 5:..ic-Pe".-

.. - : " lution c:r-centration.

m. ;l: a. p.. a r, p s m _. P :.. _ g :a ngn Ef; valves and e:.ti. ate ;:.o'.

.ec.;;iatu:_:.

Currently SNL is planning (under Tech Assistance contract) to use c;"0NA (a dD code) for transier. analyses.

It would appear, on the basis of g dis-cussions with Fuat Ocer, that any ATNS audit calculations using RA"'NA cannot be cor.pleted until sore tir.e next year ( " arch).

Since we hope to propose to the Cormission a recen. ended course of action on B'."Rs in the next few (3-??) months, I see a need for the folic.<ing:

1736 234

3. "

':.e- ***:

j ?.i.;

n ~t < l d*'

^ ' ' ' bup,aT1

,-&[,Iq Cc ;;c;e Item 1 above in two ponths.

by 8 h\\\\

h.

Nala:c Item 2a abcVe in tv;c anths (EEs could be justification c#

t.t Earlier P,ELAP studies provide sufficient hsis for new but to be f.a tr.er c3nfir.Ted later usir.g U1'.:::A or other suitable code).

.t)g

, ' 'l -

Ca.cplete item 2b early next year but prior to AT'.:S rule being effective (guess - f*erch '80).

Completion of this task would require that a fairly

/y si~ pie contain ent model be incorporated in the code,

_AV.

ty V 5

8

'.-: nile I see a need for supporting ATWS audit calculations, I do not believe b

that a 3-D core r.: del is needed to get r' ore accurate reactivity feedback D

t "e:ts.

Before v.c sign a Tech Assistarre cor. tract using PA!'. ORA, I re-como that.you,14ike, Fuat, Dan (Fieno), and I r.ee to discuss our needs and help Cuat prepare Tech Assistance request consistent with our ATWS pl5ns

.~,. E '..' r s.

/

i. l

~ -

i_-

Ashok Thadani cc:

i A;.co:L F. O a cr.y dDdf a D-

m'y a tL ~ ~.: ns. F._nr T. S;eis 1736 235}}